INTRODUCTION

Call processing time in emergency dispatch continues to be a topic of much discussion in public safety, and in emergency medical services in particular. The most widely used definition of call processing (also known as alarm processing) is: "Alarm processing begins when the alarm is answered by the telecommunicator, includes interrogation of the caller, and ends at the beginning of emergency unit notification."¹ A call processing time standard for emergency medical calls in the United States has been codified by the National Fire Protection Association in its standard, NFPA 1221 (7.4.2). The latest version of this standard (January, 2016) states: "Emergency alarm processing for the following call types shall be completed within 90 seconds 90 percent of the time and within 120 seconds 99 percent of the time: (1) Calls requiring Emergency Medical Dispatch questioning and pre-arrival instructions."² Yet a single "one-size-fits-all" time standard may be insufficient, since it does not account for the specific type of event (Chief Complaint) and individual case circumstances—including those cases with a true need for a rapid emergency medical services (EMS) response.^{3, 4} Many EMS systems include dispatch call processing time as part of their reported EMS unit response time to the scene, affecting response time performance statistics for the EMS response agency. This in turn prompts those response agencies to pressure the emergency dispatch center to shorten call processing times—in most cases without evaluating the efficacy of the entire calltaking process—including how much time is actually necessary for the Emergency Medical Dispatcher (EMD) to collect and transmit critical case information with a high degree of accuracy and completeness. Further, with the exception of the relatively few cases of sudden cardiac (or respiratory) arrest, there exists very little evidence that incrementally shorter EMS response times actually improve patient outcomes,^{5, 6, 7, 8, 9} so there appears to be little reason not to allow EMDs to complete a thorough evaluation of the patient and emergency situation before notifying response crews.

Call prioritization time (CPT) is a key sub-component of call processing time, as it includes the gathering of the patient's Chief Complaint and primary symptoms and scene safety and hazard information, as well as assigning the correct priority level and determinant coding. In the Medical Priority Dispatch System (MPDS®), the CPT begins after address and phone number verification, at the start of the Case Entry query, "Okay, tell me exactly what happened," and ends once the MPDS determinant code has been assigned to the case (see Figure 1). EMS agencies whose dispatch centers use the MPDS software—ProQA®—can measure CPT consistently, making it readily available for study. ProQA is an expert system that directs the calltaker through a structured questioning sequence, obtaining case information from the caller (including the nature of the incident and critical incident and patient details), and recording those answers in the software program as each answer is provided by the caller. Once the most important information has been obtained, ProQA's logic and domain rules recommend a dispatch determinant code that can be immediately assigned to the case by the EMD, or in rare instances, the EMD can choose to override the recommended code to assign a higher priority coding.

Medical emergencies reported to 911 vary widely in the nature, urgency, and information required to effect a specific dispatch of the correct mix of responder units. There are 33 standard Chief Complaint Protocols in the MPDS (version 12.2, 2012) (see Figure 2), with several others that are used optionally for specialized calls (e.g. medical facility calls, automatic crash notification (ACN), suspected cases of novel flu virus, etc.). Some of these Chief Complaints, such as seizures, require a somewhat longer questioning sequence in order to arrive at the correct prioritization, as compared to others such as cardiac or respiratory arrest—conditions that can be determined in the first several questions for most patients. And typically, the more urgent the need for an immediate EMS response, the shorter the CPT, since high-acuity symptoms (simply referred to as priority symptoms in MPDS parlance) will typically be asked and recorded earlier in the MPDS questioning sequences.

There are six priority levels (see Figure 3) in the MPDS. These levels are used to define the relative urgency and response needs of the patient. There are also several hundred determinant codes that further define the nature of the incident and provide more specific response assignments. Each priority level is associated with a response value or mode: COLD or HOT. The COLD response mode comprises the OMEGA, ALPHA, and CHARLIE-level calls. However, the HOT response mode comprises the BRAVO, DELTA, and ECHO-level calls.

The purpose of this study, therefore, was to determine CPT as a function of MPDS Chief Complaint type and priority level. We hypothesized that the CPT would vary based on the nature of the Chief Complaint and the priority level assigned by the EMD.

Objectives

The primary objective of this study was to determine median CPT by MPDS priority level and Chief Complaint type. A secondary objective was to quantify and qualify any differences in CPT as associated with the priority level and Chief Complaint.

METHODS

Design and Setting

This retrospective study included six emergency communication agencies, each accredited by the International Academies of Emergency Dispatch (IAED) as Emergency Medical Dispatch Centers of Excellence: Austin-Travis County EMS, Austin, TX, USA; Emergency Medical Services Authority (EMSA), Oklahoma City and Tulsa, OK, USA; Louisville Metro EMS/MetroSafe (LMEMS), Louisville, KY, USA; Mecklenburg EMS (MEDIC), Charlotte, NC, USA; MedStar Mobile Healthcare, Fort Worth, TX, USA; Salt Lake City Fire Department/SLC 911 (SLCFD), Salt Lake City, UT, USA. The study was approved by the International Academies of Emergency Dispatch (IAED) Institutional Review Board.

Study population

The study sample included all the available emergency medical calls (during the study period from January 2006 to December 2014) that were handled using the MPDS software (ProQA®): Austin EMS (11/09/2009 - 04/14/2014 [52 months]), EMSA (04/01/2004 - 12/31/2014 [129 months]), LMEMS (07/07/2009 - 10/08/2014 [62 months]), MEDIC (09/01/2011 – 01/26/2015 [41 months]), MedStar (01/10/2006 – 04/05/2014 [99 months]), and SLCFD (04/26/2012 – 12/31/2014 [32 months]).

Outcome measures

The primary endpoints in this study were the median CPT for each Chief Complaint and each priority level, and the percentage of cases prioritized within 15, 30, 45, 60, 70, 80 , 90, 120, and 180 seconds.

Data analysis

STATA software for Windows (StataCorp. 2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP.) was used for data analysis. All cases that had a CPT <5 seconds or dispatch time >300 seconds were excluded from analysis. Descriptive statistics such as median CPT (including percentiles), frequencies, and percentages were used to profile the distributions. Case distributions were presented by priority Chief Complaint Protocols and priority levels—for each agency and overall.

The nonparametric Pearson's chi-square median test was used to assess whether the median CPT values for the study groups were statistically identical—at a 0.05 level of significance. Response levels were also classified as COLD (i.e., OMEGA-level, ALPHA-level, and CHARLIE-level) or HOT (i.e., BRAVO-level, DELTA-level, and ECHO-level). Comparison of median CPT values for these two study groups was also performed.

RESULTS

A total of 3,234,290 cases were collected in this study. Of these cases, 0.55% (n = 17,703) were excluded as outliers, considered to be the result of data entry errors for having a CPT <5 seconds (n = 560) or >300 seconds (n = 17,143). An additional 0.42% (n=13,618) cases were excluded due to coding errors, and 1.3% (n=40,676) as test calls, resulting in a sample size of 3,162,290 (97.8%). Of this study sample overall, 29.4% were DELTA-level, 25.5% were BRAVO-level, 21.9% were CHARLIE-level, 17.6% were ALPHA-level, 3.8% were OMEGA-level, and 1.8% were ECHO-level calls (Table 1).

In terms of median CPT, generally, all call types were prioritized in under 60 seconds median time (see Figure 4 and Table 2). Specifically, overall, ECHO-level calls were prioritized in 33 seconds, followed by BRAVO-level (43 seconds), DELTA-level (47 seconds), OMEGA-level (52 seconds), CHARLIE-level (57 seconds), and ALPHA-level calls (58 seconds). Compared to the ALPHA-level calls (the reference measure), the median time values for each of the other priority level calls were significantly different (p<0.001).

The median time for a COLD response was significantly higher than median time for a HOT response (57 seconds and 45 seconds; p<0.001, respectively).

Overall, the highest percentage of ECHO-level (35%) and BRAVO-level calls (22%) were prioritized in 15-30 seconds (see Figure 5). Most of the DELTA-level calls (25%) were prioritized in 30-45 seconds, and the OMEGA-level calls in 30-90 seconds (22%). However, most of the ALPHA-level (30%) and CHARLIE-level calls (28%) were prioritized in 60-90 seconds.

Cumulatively, over 91% of calls were prioritized in 120 seconds or less (see Figure 6). Additionally, ECHO-level calls were prioritized in 90 seconds or less 91.4% of the time, and in 120 seconds or less 95.2% of the time. 81.7% of the calls that received a (default) COLD response, and 86.5% of calls that received a (default) HOT response, were prioritized in 90 seconds or less.

Overall, calls were prioritized in a median 51 seconds across all the agencies studied (Table 2). The three Chief Complaints that recorded the shortest CPTs were: Unknown Problems (29 seconds), Traffic/Transportation Incidents (30 seconds), and Cardiac Arrest (37 seconds). The three Chief Complaints that recorded the longest CPTs included: Inaccessible Incident/Other Entrapments (81 seconds), Pregnancy/Childbirth/Miscarriage (76 seconds), and Stroke (CVA)/Transient Ischemic Attack (TIA) (73 seconds).

The Animal Bites/Attacks Chief Complaint Protocol had the longest CPT for OMEGA-level (73 seconds), and Inaccessible Incident/Other Entrapments had the longest BRAVO-level CPT (89 seconds). The Heart Problems/AICD and Stroke (CVA)/Transient Ischemic Attack (TIA) Chief Complaints had the longest CPTs in the ALPHA level (both 110 seconds), while Convulsions/Seizures had the longest CPTs at the CHARLIE level (92 seconds), and Pregnancy/Childbirth/Miscarriage had the longest CPTs at the DELTA level (82 seconds). However, the Unconscious/Fainting (Near) Chief Complaint Protocol had the longest CPT for the ECHO-level calls (46 seconds). On the other hand, the Abdominal Pain Chief Complaint Protocol had the shortest CPT for OMEGA-level (29 seconds), ALPHA-level (36 seconds), and CHARLIE-level (49 seconds) calls. The other Chief Complaint Protocols that had the shortest CPTs were: Unknown Problems for BRAVO-level (25 seconds), Diabetic Problems for DELTA-level (27 seconds), and Burns/Explosion for ECHO-level calls (12 seconds).

In each Chief Complaint Protocol, the median time for HOT responses was significantly shorter than median time for COLD responses (p<0.05), except for Back Pain, Eye Problems/Injuries, Headache, and Pregnancy/Childbirth/Miscarriage Protocols, where CPT values for COLD response were lower. However, there was no significant difference in CPT values between HOT and COLD responses for Falls, Inaccessible Incident/Other Entrapments, Traumatic Injuries (Specific), and Pandemic/Epidemic/Outbreak Chief Complaint Protocols.

A median test comparison of CPT values for the 5 most frequently used Chief Complaint Protocols (Sick Person, Breathing Problems, Falls, Traffic/Transportation Incidents, Chest Pain) showed a significant difference between CPT values for each Chief Complaint compared to overall median CPT by priority level. For example, a coding of 17 DELTA (71 seconds) was significantly higher (p <0.001) than the overall median time for DELTA (47 seconds). In other words, it appears that Chief Complaint type makes more difference than priority level alone in determining CPT.

DISCUSSION

TABLES AND GRAPHS

Figure 1. Call prioritization time as a component of EMS response time (*not to scale) Figure 2. Medical Priority Dispatch System Chief Complaint Protocols v12.2 Figure 3. International Academies of Emergency Dispatch non-linear response matrix. Table 1. Distribution of cases by agency and priority level. Figure 4. Overall median call prioritization time by priority level Table 2. Call prioritization time (in seconds) by dispatch priority level. Figure 5. Distribution of calls prioritized within individual call prioritization time interval. Figure 6. Distribution of calls prioritized by specific call prioritization time cut-off.

DISCUSSION

Since the ECHO-level code indicates the most time-critical, life-threatening conditions—such as cardiac or respiratory arrest, complete airway obstruction, respiratory failure, or a person on fire—a rapid response by the closest trained EMS crew (or rescue personnel) is necessary to provide rapid life support for ECHO patients. And since a short response time could make a difference for these patients, ideally it is these ECHO cases that get dispatched the fastest. Our results suggest that this is indeed what is occurring. Further, the data show the DELTA-level cases—which are the next highest in urgency and response need, just below ECHO, and represent potentially life-threatening conditions—appropriately demonstrate a relatively short (but longer than ECHO) CPT.

Interestingly, the BRAVO-coded cases show a somewhat shorter time than DELTA (43 seconds and 47 seconds respectively, [p<0.001]). This is likely due to the high number of BRAVO codes that are assigned to Protocol 29 (Traffic/Transportation Incidents) and Protocol 32 (Unknown Problems/Person Down). For both of these protocols, a special 'fast-track' software feature allows the EMD to skip several questions in the Case Entry sequence under certain conditions (e.g., multiple patients, or reporting party is not at the scene), move straight to a Key Question sequence, and subsequently obtain the assigned determinant code. One further fact is worth noting in regards to the BRAVO-level: the MPDS Response Determinant matrix (Figure 2) recommends a lights-and-siren (HOT) response for BRAVO calls, just as it does for DELTAs and ECHOs—but not for CHARLIE, ALPHA, and OMEGA cases. The BRAVO-level, therefore, carries a response urgency more in line with the higher priority calls. This translates to a shorter response time by EMS crews, in many cases. The rationale behind a rapid BRAVO response is that, while these cases are generally not life-threatening, the true patient status—including treatment and transport need—is somewhat less predictable than those of the other priority levels. A rapid response by the closest available response unit (typically BLS) allows for the first responders to more rapidly assess the patient and order an upgraded or downgraded ambulance response when necessary.

CPT also varied substantially by Chief Complaint Protocol (call nature), even within the same priority level. For example, in the DELTA level, Protocol 24 (Pregnancy/Childbirth/Miscarriage) had a median CPT of 82 seconds, while the median for a DELTA-level in Protocol 13 (Diabetic Problems) was 27 seconds. These wide time differences showed up in all of the priority levels (Table 2). In fact, it is apparent from our results that the Chief Complaint Protocol used is the single most influential factor on CPT; the median time differences between the priority levels are small in comparison. The results of our study do not support either of the NFPA 1221 percentile time requirements—90% within 90 seconds, and 99% within 120 seconds. In fact, given that the elapsed time to achieve address and phone number verification is not included in our data, it would be unrealistic to expect the vast majority of Emergency Medical Dispatch agencies to meet such requirements, particularly for agencies that complete the EMD call prioritization process before notifying EMS response crews. Given the complex nature and variety of the medical Chief Complaints presented here, it is clearly necessary, and often critical, for the EMD to gather complete and accurate information at the time of, and, in most cases preceding, unit notification, which calls into question the absolute necessity of placing a firm time ceiling on this heterogeneous information gathering process.¹⁰

Finally, given our results showing wide variations in times by Chief Complaint Protocol, any time standard applied to an individual dispatch agency must necessarily account for the mix of Chief Complaint types handled by that agency. For example, if agency X had double the number of DELTA-level Pregnancy/Childbirth/Miscarriage calls (82 seconds median CPT) than agency Y—which in turn had twice the number of DELTA-level Diabetic Problems cases (27 seconds median CPT)—Agency Y would clearly have a shorter overall CPT for DELTA-level calls, assuming the mix of other DELTA-level calls was roughly equal between the two agencies. Regional and local differences in demographics, traffic patterns, and even air quality could change the mix of Chief Complaint types encountered by a dispatch agency. Indeed, given the wide variations in CPT by Chief Complaint type, as well as the very demonstrable differences in the Chief Complaint mix between 911 jurisdictions, a one-size-fits all nationwide time standard appears not only unrealistic, but self-defeating.

Limitations

As was noted in the introduction, and illustrated in Figure 1, this study does not evaluate the entire call processing time according to the widely accepted definition. The CPT is a portion of the overall call processing time and should be viewed strictly as means of understanding the medical and situational assessments that an EMD performs after the address and phone number have been verified. Future studies are necessary to examine the complete call processing time, i.e., the elapsed time from phone pickup to response unit notification.

CONCLUSION

EMD call prioritization time is a measurable component of call processing time. Median times vary significantly by both Chief Complaint (call nature) and assigned priority level. The most time-critical cases, or ECHO-level calls, had the shortest median prioritization time, followed by BRAVO- and DELTA-level calls. Wide variations in median times are seen between the various Chief Complaint Protocols. At the DELTA level, the Pregnancy/Childbirth/Miscarriage Protocol (Protocol 24) had the longest median time (82 seconds), while the Diabetic Problems Protocol (Protocol 13) had the shortest (27 seconds). A "one-size-fits-all" time standard is insufficient to govern the wide spectrum of Chief Complaint types and priority levels handled by Emergency Medical Dispatch agencies. The NFPA standard requiring 90% of call processing to be completed in 90 seconds and 99% in 120 seconds is not supported by this study's results.

ACKNOWLEDGMENTS

The study investigators wish to thank Brandee Rowley, Administrative Assistant, IAED, for coordinating the study and manuscript logistics and tracking; Jay Dornseif, Fire Protocols Consultant, Priority Dispatch Corp., for the initial coordination of data requests to the study sites; and to study sites for providing the data.

REFERENCES

1. NFPA 1221, Standard for the Installation, Maintenance, and Use of Emergency Services Communications Systems. Section A7.4.1 National Fire Protection Association. January 2013 view
2. NFPA 1221, Standard for the Installation, Maintenance, and Use of Emergency Services Communications Systems. Section 7.4.2 National Fire Protection Association. January 2016.
3. Clawson JJ, Sinclair B. Creating Time Standards That Work. The National EMD Journal. Spring 1999: 1 & 14- 15.
4. Salvucci A, Kuehl A, Clawson J, Martin R. The Response Time Myth: Does Time Matter in Responding to Emergencies? Topics in Emergency Medicine. 2004:26:2.
5. Blackwell TH, Kline JA, Willis JJ, Hicks M. Lack of Association Between Prehospital Response Times and Patient Outcomes. Prehospital Emergency Care. 2009; 13; 444-450
6. Pons PT, Markovchick VJ. Eight Minutes or Less; Does the Ambulance Response Time Guideline Impact Trauma Patient Outcome? J Emerg Med. 2002; 23:43-8
7. Pons PT, Haukoos JS, Bludworth W, Cribley T, Pons KA, Markovchick VJ. Paramedic Response Time: Does it Affect Patient Survival? Academic Emergency Medicine. 2005; 12:594–600.
8. Blanchard IE, Doig CJ, Hagel, BE, Anton AR, Augun, DA, Kortbeek JB, Powel DG, Wiliamson TS, Fick GH, Innes GD. Emergency Medical Services Response Time and Mortality in an Urban Setting. Prehospital Emergency Care. 2012; Early Online: 1-10
9. Salvucci A, Kuehl A, Clawson J, Martin R. The Response Time Myth: Does Time Matter in Responding to Emergencies? Topics in Emergency Medicine. 2004:26:2.
10. Frazier, A. Clock is Ticking, NFPA 60-second time standard rattles communications. The Journal of Emergency Dispatch. Sep/Oct 2015: 38 – 40.